December 2015
Volume 56, Issue 13
Free
Cornea  |   December 2015
Human Corneal MicroRNA Expression Profile in Fungal Keratitis
Author Affiliations & Notes
  • Hemadevi Boomiraj
    Department of Microbiology Aravind Medical Research Foundation, Madurai, India
  • Vidyarani Mohankumar
    Department of Microbiology Aravind Medical Research Foundation, Madurai, India
  • Prajna Lalitha
    Department of Microbiology Aravind Medical Research Foundation, Madurai, India
  • Bharanidharan Devarajan
    Department of Bioinformatics, Aravind Medical Research Foundation, Madurai, India
  • Correspondence: Bharanidharan Devarajan, Department of Bioinformatics, Aravind Medical Research Foundation, 1, Anna Nagar, Madurai, India; bharanid@gmail.com
Investigative Ophthalmology & Visual Science December 2015, Vol.56, 7939-7946. doi:10.1167/iovs.15-17619
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Hemadevi Boomiraj, Vidyarani Mohankumar, Prajna Lalitha, Bharanidharan Devarajan; Human Corneal MicroRNA Expression Profile in Fungal Keratitis. Invest. Ophthalmol. Vis. Sci. 2015;56(13):7939-7946. doi: 10.1167/iovs.15-17619.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose: MicroRNAs (miRNAs) are small, stable, noncoding RNA molecules with regulatory function and marked tissue specificity that posttranscriptionally regulate gene expression. However, their role in fungal keratitis remains unknown. The purpose of this study was to identify the miRNA profile and its regulatory role in fungal keratitis.

Methods: Normal donor (n = 3) and fungal keratitis (n = 5) corneas were pooled separately, and small RNA deep sequencing was performed using a sequencing platform. A bioinformatics approach was applied to identify differentially-expressed miRNAs and their targets, and select miRNAs were validated by real-time quantitative PCR (qPCR). The regulatory functions of miRNAs were predicted by combining miRNA target genes and pathway analysis. The mRNA expression levels of select target genes were further analyzed by qPCR.

Results: By deep sequencing, 75 miRNAs were identified as differentially expressed with fold change greater than 2 and probability score greater than 0.9 in fungal keratitis corneas. The highly dysregulated miRNAs (miR-511-5p, miR-142-3p, miR-155-5p, and miR-451a) may regulate wound healing as they were predicted to specifically target wound inflammatory genes. Moreover, the increased expression of miR-451a in keratitis correlated with reduced expression of its target, macrophage migration inhibitory factor, suggesting possible regulatory functions.

Conclusions: This is, to our knowledge, the first report on comprehensive human corneal miRNA expression profile in fungal keratitis. Several miRNAs with high expression in fungal keratitis point toward their potential role in regulation of pathogenesis. Further insights in understanding their role in corneal wound inflammation may help design new therapeutic strategies.

Corneal disease is a major public health problem and a leading cause of blindness, second to cataracts, in developing countries.1 In India, corneal lesions account for 9% of all blindness based on a survey by the Government of India.2 Although a healthy cornea is inherently resistant to infection, the slightest trauma can facilitate the entry of pathogens into the corneal stroma and cause keratitis.3 Fungal keratitis is characterized by rapid disease progression with corneal ulceration and stromal inflammatory infiltrate.4 The polyenes, natamycin and amphotericin B, are the mainstay of treatment in fungal keratitis, but these drugs have a limited corneal penetrance and also are ineffective during later stages of the disease.5 Nearly 15% to 27% of patients with fungal keratitis require surgical intervention, which in turn has a relatively poor prognosis.6 Clearly, there is a need for improved therapeutic approaches that can mitigate inflammation and expedite corneal wound repair in fungal keratitis. 
MicroRNAs are small noncoding RNAs (∼22 nucleotides) that regulate gene expression by complementary binding to the 3′UTR region of target mRNAs. These highly conserved small RNAs have been shown to regulate cell growth, apoptosis, and other cellular processes in eukaryotes.7 To date, more than 2600 miRNAs have been identified in the human genome (miRBase; http://www.mirbase.org/, in the public domain), and over a third of all protein-coding genes are thought to be regulated by a relatively small number of miRNAs.7 MicroRNAs are expressed in various ocular tissues with a distinct tissue-specific and developmental stage-specific expression pattern, which suggests potential unique functions.8 They have also been detected in ocular fluids9 and have been implicated in human corneal diseases like keratoconus,10 Fuchs' dystrophy,11 and herpetic stromal keratitis.12 
Corneal inflammation in fungal keratitis poses a significant therapeutic challenge to clinicians, particularly because of the controversial role of corticosteroids in the initial disease management. More recently, therapeutic modulation of miRNA expression has evolved as a promising new approach to treat human inflammatory disorders.13,14 Indeed, miRNAs have been shown to target cytokines and proteins involved in inflammatory signaling, immune cell differentiation, and host immune responses to infection.15,16 Thus, we hypothesized that dysregulations in miRNA expression could be associated with excessive corneal inflammation and impaired wound healing in fungal keratitis. In the current study, we intended to determine the human miRNA expression profile by deep sequencing the small RNA transcriptome in fungal keratitis and normal human donor corneas. To better understand their regulatory role in disease pathogenesis, we applied a bioinformatics approach to construct a miRNA target gene regulation network. 
Materials and Methods
Samples
A total of 11 posttransplant corneas were obtained during surgery from keratitis patients (Supplementary Table S1) who were culture positive for Aspergillus flavus (confirmed by sequencing the internal transcribed spacer [ITS] region). The study protocol was approved by the Institutional Review Board of Aravind Medical Research Foundation and an informed consent was obtained from each patient before surgery. The exclusion criteria were acute or chronic systemic illness and topical steroid therapy or any other form of immunosuppression. The study was performed according to the tenets of the Declaration of Helsinki. Human donor eyes were acquired from Rotary Aravind international eye bank (Madurai, India) and the corneas were trephined immediately. Corneal tissues were snap-frozen in liquid nitrogen and stored at −80°C. 
Extraction of RNA
Normal donor corneas (n = 3, mean age ± SD: 91.6 ± 1.5 years; male-to-female ratio = 2:1) and fungal keratitis corneas (n = 5, age ± SD: 54.6 ± 13.3 years; male-to-female ratio = 4:1) were pooled separately (Supplementary Table S1) to meet the input requirements of cDNA library construction and deep sequencing. Total RNA of 3 μg and 3.2 μg from normal and infected cornea pools was extracted using a miRNA isolation kit (mirVana; Life Technologies, Carlsbad, CA, USA). Briefly, the frozen corneal tissue was ground with liquid nitrogen, followed by the kit RNA extraction protocol for frozen and extremely hard tissues. Samples with a RNA integrity number greater than 8.0, as evaluated using an RNA nano chip in a bioanalyzer (RNA 6000 Nano Kit; Agilent Technologies, Waldbronn, Germany) were used for library preparation. 
Library Construction of cDNA and Deep Sequencing
Small RNA libraries were constructed using a small RNA sample preparation kit (TruSeq; Illumina, San Diego, CA) following the manufacturer's protocol. Briefly, 3′ and 5′ adapters were ligated to small RNA, followed by reverse transcription, PCR amplification with primers containing index sequences specific for each sample, and purification from 6% polyacrylamide gel of 147- to 157-bp products from pooled indexes. Libraries were validated using a DNA chip on a bioanalyzer (DNA 1000; Agilent Technologies). Cluster generation using a cluster kit (TruSeq SR Cluster Kit v3-cBot-HS, catalog no. GD-401-3001; Illumina) and sequencing using a sequencing kit (TruSeq SBS Kit v3-HS, catalog no. PE-401-3002; Illumina) on an Illumina sequencing platform (HiSeq1000 platform) was performed at the Centre for Cellular and Molecular Platforms (Bangalore, India). 
Deep Sequencing Data Analysis
Raw deep sequencing data were obtained in the FASTQ format and a quality assessment was performed with the FastQC tool. Adapter and low-quality reads were discarded using our Perl script, allowing no mismatches for adapter identification (Fig. 1). The sequencing data were further aligned to Homo sapiens hg19 genome reference, allowing for one mismatch using bowtie1 aligner17 in the sRNAbench tool.18 Next, this tool identified the validated miRNAs from miRBase (release 18).19 To predict novel miRNAs, we used a machine learning algorithm in the sRNAbench tool, allowing no mismatches with other default parameters. The miRBase database was used for the miRNA current naming conversion where we followed the −3p or −5p nomenclature. 
Figure 1
 
Analysis pipeline for the small RNA sequencing data.
Figure 1
 
Analysis pipeline for the small RNA sequencing data.
For differential expression (DE) analysis, miRNAs with the same isoMIRs were considered together, and less-represented miRNAs (0.01% of the total reads) were eliminated from analysis. Furthermore, the analysis was performed using the NOISeq R software package on the quantile normalized reads (Bioconductor 2.14 with R version 3.1.0). The method of NOISeq applied was NOISeqsim, which allows for DE testing with no replicates. The variable expression of miRNAs between patients and controls was considered significant when the fold change was 2 or greater and the probability score was more than 0.9. The evaluation of the results, represented by a volcano plot in Figure 2, was based on the NOISeq scores: the M-value and D-value. 
Figure 2
 
Volcano plot showing differentially expressed miRNAs. Red dots represent significant differentially-expressed miRNAs, and black dots are M- and D-values in noise. M-value: log2(x1/x2). D-value: |x1x2|, where x1 and x2 are the expression levels in keratitis and donor cornea, respectively. Arrows mark the validated miRNAs.
Figure 2
 
Volcano plot showing differentially expressed miRNAs. Red dots represent significant differentially-expressed miRNAs, and black dots are M- and D-values in noise. M-value: log2(x1/x2). D-value: |x1x2|, where x1 and x2 are the expression levels in keratitis and donor cornea, respectively. Arrows mark the validated miRNAs.
MicroRNA Target Gene Prediction and Functional Analysis
The best predicted target genes for each significant differentially expressed miRNA were identified using four different algorithms: DIANA-microT,20 TargetScan,21 miRanda,22 and PITA.23 The targets that represent the intersection of at least three algorithms were selected to avoid false positives. Together with these targets, the experimentally validated targets available in the Tarbase24 and mirRecords25 database were included. The target genes were further filtered as putative target genes based on the corneal proteome database that was created from reported human corneal proteins under normal and any corneal disease conditions. In order to predict the functions of these miRNAs in the regulation of disease pathogenesis, we performed the functional analysis of putative target genes using the Database for Annotation, Visualization, and Integrated Discovery (DAVID).26 The putative target genes were grouped into functional categories using the Gene Ontology (GO) term and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways. The P value less than 0.01 and false discovery rate (FDR) less than 0.01 were used as the threshold to select significant GO categories and KEGG pathways. 
Cytoscape (version 3.0.1) software was used to construct the miRNA target gene network and was enriched with GO categories of interest. For the topologic analysis, the closeness centrality was used that calculates how many steps a specific node requires to connect to other nodes. 
Real-Time Quantitative PCR
To validate the sequencing results, quantitative PCR (qPCR) was done with RNA obtained from six patient corneas (mean age ± SD: 56.1 ± 13.1 years; male-to-female ratio: 3:3) and six donor corneas (mean age ± SD: 80.6 ± 9.8 years; male-to-female ratio: 3:3) (Supplementary Table S1). Mature microRNAs were polyadenylated and reverse transcribed using an RT kit (miScript II RT Kit; Qiagen, Valencia, CA, USA) whereby a universal tag is incorporated at the 5′ end of the cDNA. This tag is bound by the universal reverse primer during real-time PCR amplification (miScript SYBR Green PCR Kit; Qiagen) with custom synthesized forward primers (Supplementary Table S2). The reaction conditions included an initial activation step at 95°C for 15 minutes, followed by 40 cycles of 94°C for 15 seconds, 55°C for 30 seconds, and 70°C for 40 seconds. Data were normalized using RNU6B (U6) snRNA levels. 
For quantitative analysis of mRNA expression, cDNA was synthesized using a high-capacity cDNA reverse transcription kit (Applied Biosystems, Carlsbad, CA, USA), and qPCR was done with a PCR master mix (Power SYBR Green; Applied Biosystems) and custom-made primers (Supplementary Table S2). The PCR amplification was carried out at 50°C for 2 minutes, 95°C for 10 minutes, and 40 cycles of 95°C for 15 seconds and 60°C for 1 minute with a final melt curve analysis. The reactions were run in triplicate and the results were normalized using β-actin mRNA levels. The relative expression between control and patient samples was calculated by the 2^−ΔΔCt method. A Mann-Whitney U test was used for the statistical analysis of qPCR data. P less than 0.05 was considered as statistically significant. 
Data Access
The small RNA sequence data was submitted to the Gene Expression Omnibus (https://www.ncbi.nlm.nih.gov/projects/geo/, in the public domain) under accession number GSE64843. Novel miRNA sequences have been submitted to miRBase for assignment of an accession number. 
Results
Characteristics of Deep Sequencing and miRNA Composition
To determine the miRNA expression profile in fungal keratitis, small RNA sequencing was performed with independently pooled RNA samples from donor and fungal keratitis corneas. More than 20 million raw reads were generated for each pooled sample and approximately 98% raw reads mapped to human genome hg19 using the sRNAbench tool (Fig. 1). Only 27% and 34% of reads were mapped to mature miRNAs with miRBase (release 18), which identified 1752 and 1527 mature miRNAs from keratitis and donor corneal samples, respectively. To consider the most robust miRNA expression, miRNAs with 0.01% of total reads were eliminated in both normal and infected samples, which returned 544 miRNAs for further analysis. An initial insight into the miRNA expression profile (Supplementary Table S3) highlights a dynamic modulation of large number of mature miRNAs during infection. 
To evaluate the significance of DE by the 544 miRNAs, we calculated the probability of being differentially expressed using NOISeq scores, M-value and D-value. Based on these calculations, 75 miRNAs showed significant DE in keratitis with a fold change 2 or greater. Of these, 43 miRNAs were upregulated (those with a positive M-value) and 32 were downregulated (those with a negative M-value) as summarized in the volcano plot (Fig. 2). The miR-21-5p, miR-146b-5p, miR-143-3p, miR-204-5p, and miR-184 had relatively more abundant expressions in keratitis and control corneas with greater than 2-fold DE (Supplementary Table S4; Supplementary Fig. S1). 
Validation of Differentially Expressed miRNAs by Real-Time PCR
Based on high DE (≥6-fold) and functional relevance to corneal diseases, 16 miRNAs were selected for further validation by qPCR (Table 1). The clinical details, ulcer characteristics, and final visual acuity of keratitis patients are summarized in Supplementary Table S1. In general, the qPCR data were in accordance with the deep sequencing data, although the magnitude of fold difference was higher with qPCR (Table 1; Figs. 3A, 3B). In keratitis, the highest level of expression (1530-fold by qPCR) was noted for miR-223-3p, which is known to regulate cell proliferation, differentiation, and wound inflammation.27,28 Since the amount of gene expression correlates inversely with the ΔCt values, the normalized ΔCt levels were compared between donor and keratitis corneas. While all the upregulated miRNAs had significantly lower ΔCt values in keratitis (Supplementary Figs. S2A, S2B), the downregulated miRNAs, miR-204-5p (P = 0.026) and miR-184 (P = 0.043) had significantly lower ΔCt values in controls (Supplementary Fig. S2C). 
Table 1
 
Differentially-Expressed miRNAs Validated Using Real-Time qPCR
Table 1
 
Differentially-Expressed miRNAs Validated Using Real-Time qPCR
Figure 3
 
Relative expression of miRNAs and target genes in fungal keratitis. Real-time PCR validation of differentially expressed miRNAs and select target genes in fungal keratitis (n = 6) and donor corneas (n = 6). Relative expression levels of (A) upregulated miRNAs, (B) downregulated miRNAs, and (C) MIF, CXCR2, and CXCR4 genes are shown. Data are expressed as the log of relative fold difference. Each data point represents individual patient sample.
Figure 3
 
Relative expression of miRNAs and target genes in fungal keratitis. Real-time PCR validation of differentially expressed miRNAs and select target genes in fungal keratitis (n = 6) and donor corneas (n = 6). Relative expression levels of (A) upregulated miRNAs, (B) downregulated miRNAs, and (C) MIF, CXCR2, and CXCR4 genes are shown. Data are expressed as the log of relative fold difference. Each data point represents individual patient sample.
Target Genes and Functional Analysis
To explore the various targets and their associated pathways, a putative target list was obtained (see Methods; Supplementary Table S4). For functional analysis, the DAVID database was used to integrate the target genes into common pathways or GO process. In total, 270 of 590 genes submitted were annotated to 55 KEGG pathways, and 38 pathways were significantly regulated (P < 0.01). The highly regulated pathways with FDR less than 0.01 are shown in Table 2. Most of these pathways, including focal adhesions, toll-like receptors (TLRs), cancer-related pathways, and neurotrophin-signaling pathways, regulate cell migration, growth, proliferation, angiogenesis, and inflammation,16,29,30 which indicates strong regulatory roles for miRNAs in corneal wound healing.31 
Table 2
 
Significant Pathways Predicted to Be Highly Regulated by Differentially-Expressed miRNAs
Table 2
 
Significant Pathways Predicted to Be Highly Regulated by Differentially-Expressed miRNAs
MicroRNA Target Gene Regulation Network
To better understand the role of miRNAs in corneal inflammation and wound healing, we constructed a miRNA target gene regulation network. More than 46 differentially expressed significant miRNAs were enriched in inflammatory- and wound-healing–related GO categories (Supplementary Table S5). The target genes annotated with higher closeness centrality (red circle, Fig. 4) were those involved in both GO categories reported to be involved in wound inflammation, which is the first event that determines the fate and quality of healing.28 All the highly expressed miRNAs (>6-fold; darker blue) in keratitis regulate inflammatory and immune responses. Of these, miR-21-5p, miR-618, miR-144-3p, and miR-146b-5p include wound inflammatory genes as some of their targets, while miR-511-5p, miR-451a, miR-142-3p, and miR-155-5p target only wound inflammatory genes. The highly downregulated miR-124-3p targets ANXA8, CCL2, CD59, GSN, MTPN, and STAT3 genes involved in wound inflammation and healing. 
Figure 4
 
The miRNA target gene regulation network for the functional GO terms of inflammatory and immune responses and wound healing and responses. Target genes having higher centrality were marked with a red circle. The intensity of the triangle color corresponds to the level of fold change in fungal keratitis. Red: downregulated miRNAs; blue: upregulated miRNAs.
Figure 4
 
The miRNA target gene regulation network for the functional GO terms of inflammatory and immune responses and wound healing and responses. Target genes having higher centrality were marked with a red circle. The intensity of the triangle color corresponds to the level of fold change in fungal keratitis. Red: downregulated miRNAs; blue: upregulated miRNAs.
Of interest for wound inflammation, miR-511-5p targets TLR4, miR-142-3p targets RAC1, and miR-155-5p targets MYD88, TNF, and CEBPB involved in the TLR-signaling pathway, which is one of the highly regulated pathways in keratitis (Table 2). Interestingly, macrophage migration inhibitory factor (MIF), the putative target of miR-451a, has been reported to be involved in apoptosis, inflammation, cell proliferation, and wound healing.32,33 We hypothesize here that miR-451a has a specific role in wound inflammation and could serve as a potential target for corneal epithelial wound healing. 
Expression of MIF and Its Receptors in Fungal Keratitis
The relative mRNA expression levels of MIF and its receptors CXCR2 and CXCR4 was determined by qPCR with the same set of RNA samples used for validating miRNA expression. CXCR2 and CXCR4 are the cellular receptors of MIF through which it regulates wound-closure via calcium signaling.34,35 Macrophage migration inhibitory factor expression was relatively lower in keratitis (mean 2-fold; Fig. 3C), which correlates with an increased expression of miR-451a that targets MIF. CXCR4 expression levels were variable, whereas CXCR2 was consistently upregulated in all keratitis samples (mean 30-fold; Fig. 3C) with correspondingly lower ΔCt values (P = 0.002; Supplementary Fig. S2D). Previously, we have reported an increased expression of CXCR2 ligand, IL-8, in fungal keratitis corneas.36 
Novel miRNA
One bona fide novel miRNA matching with human genome was detected and did not show orthologs. The reads for pre-miRs and two arms were detected only in the keratitis corneas. The novel miRNAs (named miR-cornea-5p and miR-cornea-3p) are located in the potential stem-loop region with the predicted free energy of −39.10 kcal/mol (Fig. 5A). When validated by qPCR, the 5p and 3p species were expressed in both donor and keratitis corneas with 8.4-fold and 13.9-fold higher expressions, respectively, in keratitis, suggesting a role in disease pathogenesis (Fig. 5B). Both novel miRNAs are positioned in chromosome 8 within the coding region of early growth response 3 (EGR3) gene. In order to predict EGR3 as a potential target, DIANA-microT3.0 server was used. The miRNA miR-cornea-3p was predicted to target EGR3 with a precision binding score of 1.1. EGR3 regulates the expression of nearly 330 genes, many of which are involved in immune responses and inflammatory processes.37 EGR3 mRNA expression was downregulated in half of the patient samples, whereas in the other half it was not much different from that of controls (Fig. 5B). 
Figure 5
 
Bona fide novel microRNA. (A) Predicted secondary stem-loop structure of pre–miR-cornea and sequence read of 3p and 5p species are shown. (B). Relative expression levels of miR-cornea-3p, miR-cornea-5p, and the target gene EGR3 in keratitis (n = 6) compared to donor corneas (n = 6). Data are expressed as the log of relative fold difference, and each data point represents individual patient sample.
Figure 5
 
Bona fide novel microRNA. (A) Predicted secondary stem-loop structure of pre–miR-cornea and sequence read of 3p and 5p species are shown. (B). Relative expression levels of miR-cornea-3p, miR-cornea-5p, and the target gene EGR3 in keratitis (n = 6) compared to donor corneas (n = 6). Data are expressed as the log of relative fold difference, and each data point represents individual patient sample.
Discussion
Keratitis is an inflammatory disease that can progress rapidly with corneal ulceration through pathologic wound healing.4 Corneal infection by filamentous fungus is accompanied by a localized inflammatory response with the secretion of pro-inflammatory cytokines and recruitment of neutrophils to the site of infection.36 Normally, this is followed by an active wound-healing phase, when the inflammation resolves and the cornea undergoes fibrotic scarring. However, in many patients the inflammation does not resolve completely, and this culminates in corneal opacification and loss of vision.38 Recent reports on altered miRNA expression in human corneal diseases1012 suggest their regulatory role in pathogenesis. However, the miRNA expression profile in fungal keratitis has not yet been studied. Here, we identified the corneal miRNA expression profile in fungal keratitis patients versus normal donors using the small RNA deep sequencing method. We also depicted the regulation network of significant differentially expressed miRNAs and investigated their possible role in wound inflammation. Such a study could help us to better understand the miRNA regulatory role in pathogenesis and to identify new therapeutic strategies. 
Overall, several miRNAs were identified in normal and infected corneas (see results). Of these, only 75 miRNAs showed significant DE in keratitis (see volcano plot). Many corneal miRNAs not reported earlier were identified by deep sequencing, demonstrating a more highly sensitive detection than that of microarray studies.39 Importantly, several miRNAs with profound changes in expression were found in infected corneas. Notably, expression of miR-184 and miR-204-5p, which were abundant in the normal cornea (as shown in earlier studies40), was dysregulated in keratitis since they may be involved in corneal epithelial cell proliferation and migration during wound repair.40 Moreover, several less-abundant miRNAs were also dysregulated, especially miR-511, miR-451a, and miR-223-3p, with significantly altered expressions in keratitis may have a regulatory role in disease pathogenesis. This divergence in expression profiles could possibly be due to the presence of infiltrating immune cells during infection compared to a normal condition with sparse resident immune cells like macrophages and dendritic cells.41 The neutrophils that comprise 65% to 75% of the infiltrating cells in fungal-infected posttransplant corneas36 have been shown to be a predominant source for miR-223.42 Furthermore, using comprehensive next generation sequencing (NGS) data, we discovered a bona fide novel miRNA that was differentially upregulated in infected corneas. We further validated our findings by qPCR using independent corneal samples. 
Our earlier study reported that the cellular inflammatory response between A. flavus and Fusarium-infected corneas were comparable during early and late stages of the disease.36 In this regard, we attempted to perform a comparative analysis of miRNA expression (select differentially expressed miRNAs) between A. flavus and Fusarium-infected corneas by qPCR. Corneas from bullous keratopathy patients were used as noninfectious inflammatory controls. Interestingly, we found a significant variation in the magnitude of miRNA expression between the two fungal infections, especially with miR-204-5p and miR-cornea-3p (Hemadevi B, Vidyarani M, Prajna L, Bharanidharan D, unpublished data, 2015). Notably, expression of miR-223-3p was several-fold higher in fungal keratitis corneas compared to bullous keratopathy (Hemadevi B, Vidyarani M, Prajna L, Bharanidharan D, unpublished data, 2015), which further indicates the infiltrating neutrophils as a major source of this miRNA during corneal infections. 
The function of a miRNA is ultimately defined by its effect on the expression of those genes that it targets. Though putative targets of miRNAs can be achieved by various databases, the false-positive rate is unacceptably high because of poor sequence complementarity of miRNA-target interaction in humans. In this report, we used an efficient bioinformatics approach to analyze the regulatory role of differentially expressed miRNAs systematically. With this method, we could identify cornea-specific targets, thereby reducing a large number of relatively speculative targets. Based on the targets, we then predicted pathways and biological processes (using DAVID) controlling wound healing (e.g., focal adhesion, proliferation, and migration) and inflammation (e.g., TLR) that may be disrupted in keratitis. Interestingly, the neurotrophin-signaling pathway was significantly regulated and may be involved in corneal wound repair as reported previously in bronchial epithelial cells and dermal fibroblasts.43,44 
We also developed a miRNA target gene regulation network in wound healing and inflammation. Such an approach helped us to identify a subset of miRNA expression profiles and their targets in wound inflammation, an important step in wound healing.28 The validation of individual protein and miRNA changes will be required in subsequent studies to confirm their regulatory role in wound healing. For the subset of expression profiles, we predicted that the highly dysregulated miRNAs (miR-511-5p miR-142-3p, and miR-155-5p) involved in the TLR-pathway, among other pathways, may be involved in wound healing, as they target only wound inflammatory genes (Fig. 4). Specifically, we predicted a role for miR-451a in wound inflammation by targeting MIF. Macrophage migration inhibitory factor has been proven to be a potential gene target of miR-451 in cancer cells,45 where overexpression of miR-451 -downregulates mRNA and protein levels of MIF. Moreover, MIF has been shown to play a central role in wound healing by regulating both the inflammatory and proliferation/migration phases of wound healing.33 Furthermore, inhibition of MIF has been shown to reduce the consequences of bacterial keratitis in mice, suggesting that it may have therapeutic effects.46 Here, we report an inverse correlation in the expression of miR-451a and MIF in infected corneas, which may have physiological relevance in wound healing. However, further studies are required to prove the direct effect of miR-451a on downregulating corneal MIF expression at the protein level. 
In summary, we report for the first time (to the best of our knowledge) a comprehensive human corneal miRNA expression profile in fungal keratitis using a deep sequencing approach. We identified several differentially expressed miRNAs in infected corneas, which may lay groundwork for new therapeutic strategies. Furthermore, we developed a bioinformatics approach to identify a set of miRNA gene targets involved in the control of inflammatory responses as well as wound healing. Our work indicates that miRNAs play important regulatory roles in corneal wound inflammation, and specifically, miR-451a can be considered a potential target for further investigation. 
Acknowledgments
The authors thank the Cornea Clinic, Aravind Eye Hospital, Madurai, India, for the samples. 
Supported by Aravind Medical Research Foundation. Hemadevi Boomiraj is supported by a research associate fellowship provided by Council of Scientific and Industrial Research, Government of India. 
Disclosure: H. Boomiraj, None; V. Mohankumar, None; P. Lalitha, None; B. Devarajan, None 
References
Whitcher JP, Srinivasan M, Upadhyay MP. Corneal blindness: a global perspective. Bull World Health Organ. 2001; 79: 214–221.
National Survey on Blindness. 1991–2001. Report 2002. New Delhi, India: Ministry of Health and Family Welfare, 2002.
Garg P, Rao GN. Corneal ulcer: diagnosis and management. Community Eye Health. 1999; 12: 21–23.
Thomas PA. Fungal infections of the cornea. Eye. 2003; 17: 852–862.
Srinivasan M. Fungal keratitis. Curr Opin Ophthalmol. 2004; 15: 321–327.
Vemuganti GK, Garg P, Gopinathan U, et al. Evaluation of agent and host factors in progression of mycotic keratitis: a histologic and microbiologic study of 167 corneal buttons. Ophthalmology. 2002; 109: 1538–1546.
Valinezhad Orang A, Safaralizadeh R, Kazemzadeh-Bavili M. Mechanisms of miRNA-mediated gene regulation from common downregulation to mRNA-specific upregulation. Int J Genomics. 2014; 2014: 970607.
Karali M, Peluso I, Gennarino VA, et al. miRNeye: a microRNA expression atlas of the mouse eye. BMC Genomics. 2010; 11: 715.
Tanaka Y, Tsuda S, Kunikata H, et al. Profiles of extracellular miRNAs in the aqueous humor of glaucoma patients assessed with a microarray system. Sci Rep. 2014; 4: 5089.
Hughes AE, Bradley DT, Campbell M, et al. Mutation altering the miR-184 seed region causes familial keratoconus with cataract. Am J Hum Genet. 2011; 89: 628–633.
Matthaei M, Hu J, Kallay L, et al. Endothelial cell microRNA expression in human late-onset Fuchs' dystrophy. Invest Ophthalmol Vis Sci. 2014; 55: 216–225.
Mulik S, Bhela S, Rouse BT. Potential function of miRNAs in herpetic stromal keratitis. Invest Ophthalmol Vis Sci. 2013; 54: 563–573.
van Rooij E, Kauppinen S. Development of microRNA therapeutics is coming of age. EMBO Mol Med. 2014; 6: 851–864.
Hassan T, McKiernan PJ, McElvaney NG, et al. Therapeutic modulation of miRNA for the treatment of proinflammatory lung diseases. Expert Rev Anti Infect Ther. 2012; 10: 359 –3.
Ma X, Becker Buscaglia LE, Barker JR, Li Y. MicroRNAs in NF-kappaB signaling. J Mol Cell Biol. 2011; 3: 159–166.
He X, Jing Z, Cheng G. MicroRNAs: new regulators of Toll-like receptor signaling pathways. Biomed Res Int. 2014; 2014: 945169.
Langmead B, Trapnell C, Pop M, Salzberg SL. Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome Biol. 2009; 10: R25.
Barturen G, Rueda A, Hamberg M, et al. sRNAbench: profiling of small RNAs and its sequence variants in single or multi-species high-throughput experiments. Methods Next Gener Seq. 2014; 1: 21–31.
Kozomara A. Griffiths-Jones S. miRBase: integrating microRNA annotation and deep-sequencing data. Nucleic Acids Res. 2011; 39: D152–D157.
Maragkakis M, Reczko M, Simossis VA, et al. DIANA-microT web server: elucidating microRNA functions through target prediction. Nucleic Acids Res. 2009; 37: W273–W276.
Lewis BP, Burge CB, Bartel DP. Conserved seed pairing, often flanked by adenosines, indicates that thousands of human genes are microRNA targets. Cell. 2005; 120: 15–20.
John B, Enright AJ, Aravin A, Tuschl T, Sander C, Marks DS. Human microRNA targets. PLoS Biol. 2004; 2: e363.
Kertesz M, Iovino N, Unnerstall U, Gaul U, Segal E. The role of site accessibility in microRNA target recognition. Nat Genet. 2007; 39: 1278–1284.
Sethupathy P, Corda B, Hatzigeorgiou AG. TarBase: a comprehensive database of experimentally supported animal microRNA targets. RNA. 2006; 12: 192–197.
Xiao F, Zuo Z, Cai G, Kang S, Gao X, Li T. miRecords: an integrated resource for microRNA-target interactions. Nucleic Acids Res. 2009; 37: D105–D110.
Huang DW, Sherman BT, Lempicki RA. Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists. Nucleic Acids Res. 2009; 37: 1–13.
Wei Y, Yang J, Yi L, et al. MiR-223-3p targeting SEPT6 promotes the biological behavior of prostate cancer. Sci Rep. 2014; 4: 7546.
Roy S, Sen CK. miRNA in wound inflammation and angiogenesis. Microcirculation. 2012; 19: 224 –2.
Zhao X, Guan JL. Focal adhesion kinase and its signaling pathways in cell migration and angiogenesis. Adv Drug Deliv Rev. 2011; 63: 610–615.
You L, Kruse FE, Völcker HE. Neurotrophic factors in the human cornea. Invest Ophthalmol Vis Sci. 2000; 41: 692–702.
Maycock NJR, Marshall J. Genomics of corneal wound healing: a review of the literature. Acta Ophthalmol. 2014; 92: e170–e184.
Tillmann S, Bernhagen J, Noels H. Arrest functions of the MIF ligand/receptor axes in atherogenesis. Front Immunol. 2013; 4: 115.
Gilliver SC, Emmerson E, Bernhagen J, Hardman MJ. MIF: a key player in cutaneous biology and wound healing. Exp Dermatol. 2011; 20: 1–6.
Bernhagen J, Krohn R, Lue H, et al. MIF is a noncognate ligand of CXC chemokine receptors in inflammatory and atherogenic cell recruitment. Nat Med. 2007; 13: 587–596.
Dewor M, Steffens G, Krohn R, Weber C, Baron J, Bernhagen J. Macrophage migration inhibitory factor (MIF) promotes fibroblast migration in scratch-wounded monolayers in vitro. FEBS Lett. 2007; 581: 4734–4742.
Karthikeyan RS, Leal SM, Prajna NV, et al. Expression of innate and adaptive immune mediators in human corneal tissue infected with Aspergillus or fusarium. J Infect Dis. 2011; 204: 942–950.
Baron VT, Pio R, Jia Z, Mercola D. Early growth response 3 regulates genes of inflammation and directly activates IL6 and IL8 expression in prostate cancer. Br J Cancer. 2015; 112: 755–764.
Leal SM, Pearlman E. The role of cytokines and pathogen recognition molecules in fungal keratitis – insights from human disease and animal models. Cytokine. 2012; 58: 107–111.
Funari VA, Winkler M, Brown J, Dimitrijevich SD, Ljubimov AV, Saghizadeh M. Differentially expressed wound healing-related microRNAs in the human diabetic cornea. PLoS One. 2013; 8: e84425.
An J, Chen X, Chen W, et al. MicroRNA expression profile and the role of miR-204 in corneal wound healing. Invest Ophthalmol Vis Sci. 2015; 56: 3673.
Hamrah P, Huq SO, Liu Y, Zhang Q, Dana MR. Corneal immunity is mediated by heterogeneous population of antigen-presenting cells. J Leukoc Biol. 2003; 74: 172–178.
Allantaz F, Cheng DT, Bergauer T, et al. Expression profiling of human immune cell subsets identifies miRNA-mRNA regulatory relationships correlated with cell type specific expression. PLoS One. 2012; 7: e29979.
Szczepankiewicz A, Lackie PM, Holloway JW. Altered microRNA expression profile during epithelial wound repair in bronchial epithelial cells. BMC Pulm Med. 2013; 13: 63.
Palazzo E, Marconi A, Truzzi F, et al. Role of neurotrophins on dermal fibroblast survival and differentiation. J Cell Physiol. 2012; 227: 1017–1025.
Bandres E, Bitarte N, Arias F, et al. microRNA-451 regulates macrophage migration inhibitory factor production and proliferation of gastrointestinal cancer cells. Clin Cancer Res. 2009; 15: 2281–2290.
Gadjeva M, Nagashima J, Zaidi T, Mitchell RA, Pier GB. Inhibition of macrophage migration inhibitory factor ameliorates ocular Pseudomonas aeruginosa-induced keratitis. PLoS Pathog. 2010; 6: e1000826.
Zhang C, Chi YL, Wang PY, et al. miR-511 and miR-1297 inhibit human lung adenocarcinoma cell proliferation by targeting oncogene TRIB2. PLoS One. 2012; 7: e46090.
Zhang Z, Li Z, Li Y, Zang A. MicroRNA and signaling pathways in gastric cancer. Cancer Gene Ther. 2014; 21: 305–316.
Kästingschäfer CS, Schäfer SD, Kiesel L, Götte M. miR-142-3p is a novel regulator of cell viability and proinflammatory signalling in endometrial stroma cells. Reprod Biomed Online. 2015; 30: 553–556.
Cheng Q, Zhang X, Xu X, Lu X. MiR-618 inhibits anaplastic thyroid cancer by repressing XIAP in one ATC cell line. Ann Endocrinol (Paris). 2014; 75: 187–193.
Kee HJ, Park S, Kwon J-S, et al. B cell translocation gene, a direct target of miR-142-5p, inhibits vascular smooth muscle cell proliferation by down-regulating cell cycle progression. FEBS Lett. 2013; 587: 2385–2392.
Xu R, Bi C, Song J, et al. Upregulation of miR-142-5p in atherosclerotic plaques and regulation of oxidized low-density lipoprotein-induced apoptosis in macrophages. Mol Med Rep. 2015; 11: 3229–3234.
Hu Y-W, Hu Y-R, Zhao J-Y, et al. An agomir of miR-144-3p accelerates plaque formation through impairing reverse cholesterol transport and promoting pro-inflammatory cytokine production. PLoS One. 2014; 9: e94997.
Matsushita R, Seki N, Chiyomaru T, et al. Tumour-suppressive microRNA-144-5p directly targets CCNE1/2 as potential prognostic markers in bladder cancer. Br J Cancer. 2015; 113: 282 –28.
Figure 1
 
Analysis pipeline for the small RNA sequencing data.
Figure 1
 
Analysis pipeline for the small RNA sequencing data.
Figure 2
 
Volcano plot showing differentially expressed miRNAs. Red dots represent significant differentially-expressed miRNAs, and black dots are M- and D-values in noise. M-value: log2(x1/x2). D-value: |x1x2|, where x1 and x2 are the expression levels in keratitis and donor cornea, respectively. Arrows mark the validated miRNAs.
Figure 2
 
Volcano plot showing differentially expressed miRNAs. Red dots represent significant differentially-expressed miRNAs, and black dots are M- and D-values in noise. M-value: log2(x1/x2). D-value: |x1x2|, where x1 and x2 are the expression levels in keratitis and donor cornea, respectively. Arrows mark the validated miRNAs.
Figure 3
 
Relative expression of miRNAs and target genes in fungal keratitis. Real-time PCR validation of differentially expressed miRNAs and select target genes in fungal keratitis (n = 6) and donor corneas (n = 6). Relative expression levels of (A) upregulated miRNAs, (B) downregulated miRNAs, and (C) MIF, CXCR2, and CXCR4 genes are shown. Data are expressed as the log of relative fold difference. Each data point represents individual patient sample.
Figure 3
 
Relative expression of miRNAs and target genes in fungal keratitis. Real-time PCR validation of differentially expressed miRNAs and select target genes in fungal keratitis (n = 6) and donor corneas (n = 6). Relative expression levels of (A) upregulated miRNAs, (B) downregulated miRNAs, and (C) MIF, CXCR2, and CXCR4 genes are shown. Data are expressed as the log of relative fold difference. Each data point represents individual patient sample.
Figure 4
 
The miRNA target gene regulation network for the functional GO terms of inflammatory and immune responses and wound healing and responses. Target genes having higher centrality were marked with a red circle. The intensity of the triangle color corresponds to the level of fold change in fungal keratitis. Red: downregulated miRNAs; blue: upregulated miRNAs.
Figure 4
 
The miRNA target gene regulation network for the functional GO terms of inflammatory and immune responses and wound healing and responses. Target genes having higher centrality were marked with a red circle. The intensity of the triangle color corresponds to the level of fold change in fungal keratitis. Red: downregulated miRNAs; blue: upregulated miRNAs.
Figure 5
 
Bona fide novel microRNA. (A) Predicted secondary stem-loop structure of pre–miR-cornea and sequence read of 3p and 5p species are shown. (B). Relative expression levels of miR-cornea-3p, miR-cornea-5p, and the target gene EGR3 in keratitis (n = 6) compared to donor corneas (n = 6). Data are expressed as the log of relative fold difference, and each data point represents individual patient sample.
Figure 5
 
Bona fide novel microRNA. (A) Predicted secondary stem-loop structure of pre–miR-cornea and sequence read of 3p and 5p species are shown. (B). Relative expression levels of miR-cornea-3p, miR-cornea-5p, and the target gene EGR3 in keratitis (n = 6) compared to donor corneas (n = 6). Data are expressed as the log of relative fold difference, and each data point represents individual patient sample.
Table 1
 
Differentially-Expressed miRNAs Validated Using Real-Time qPCR
Table 1
 
Differentially-Expressed miRNAs Validated Using Real-Time qPCR
Table 2
 
Significant Pathways Predicted to Be Highly Regulated by Differentially-Expressed miRNAs
Table 2
 
Significant Pathways Predicted to Be Highly Regulated by Differentially-Expressed miRNAs
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×